AI Startup Simulator

A fun, educational, and realistic simulation that demonstrates business thinking and AI reasoning through multiple expert agents collaborating to build a startup from scratch.

Overview

This simulator uses multiple AI agents (CEO, CTO, Marketing, Investor) to collaboratively develop a startup idea, evaluate it, and produce comprehensive business documentation.

Features

• Multi-Agent Collaboration: Four specialized agents work together to build a startup
• Realistic Workflow: Simulates real startup development process
• Comprehensive Outputs: Generates pitch deck, roadmap, and market analysis
• Educational: Demonstrates business thinking and technical tradeoffs

Installation

1. Clone this repository
2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file in the root directory with your LLM configuration:

   # For Ollama (local)
   LLM_API_KEY=not-needed
   LLM_BASE_URL=http://localhost:11434
   LLM_MODEL=qwen3:8b
   
   # For OpenAI-compatible APIs
   # LLM_API_KEY=sk-...
   # LLM_BASE_URL=https://api.openai.com/v1
   # LLM_MODEL=gpt-4

Usage

Run the simulator:

python src/main.py

The simulator will:

1. CEO proposes a startup idea
2. CTO evaluates feasibility and proposes architecture
3. Marketing agent builds market analysis and GTM plan
4. Investor agent critiques the startup
5. CEO responds and refines the strategy

Outputs will be saved in the outputs/ directory.

Project Structure

ai-startup-simulator/
├── README.md
├── requirements.txt
├── .env (create this file)
├── .gitignore
├── src/
│   ├── main.py
│   ├── orchestrator.py
│   ├── utils/
│   │   ├── llm.py
│   │   ├── json_parse.py
│   │   └── render.py
│   ├── agents/
│   │   ├── ceo.py
│   │   ├── cto.py
│   │   ├── marketing.py
│   │   ├── investor.py
│   │   └── ceo_revise.py
│   └── templates/
│       ├── pitch_deck.md.j2
│       ├── roadmap.md.j2
│       └── market_analysis.md.j2
└── outputs/

Agents

CEO Agent

• Defines startup vision, problem statement, target users, and core value proposition
• Makes high-level product and business decisions
• Resolves conflicts between agents

CTO Agent

• Designs system architecture and technical stack
• Explains technical tradeoffs in simple terms
• Ensures feasibility and scalability

Marketing Agent

• Defines target market and ICP
• Builds GTM (go-to-market) strategy
• Crafts positioning and messaging

Investor Agent

• Evaluates the startup critically
• Asks tough questions about market size, moat, traction, and risks
• Provides feedback as if in a pitch meeting

Outputs

The simulator generates:

1. Pitch Deck: Problem, Solution, Product, Market Opportunity, Business Model, Traction, Competitive Advantage, Vision
2. Product Roadmap: MVP, Short-term (3-6 months), Long-term (12+ months)
3. Market Analysis: Target users, Market size, Competitors, Differentiation
4. Why This Startup Stands Out: Clear moat, AI leverage, Unique insight

Configuration

Set the following environment variables in .env:

• LLM_API_KEY: Your LLM API key
• LLM_BASE_URL: Base URL for the LLM API (e.g., for Ollama: http://localhost:11434)
• LLM_MODEL: Model name (default: qwen3:8b)

License

MIT